Drug abuse is a major global problem with often devastating societal consequences. While therapeutics designed to treat drug addiction have focused on medicinal intervention with concurrent behavioral therapy, much progress remains to be made in reducing this major social and economic burden. All drugs of abuse appear to work on the dopamine (DA) pathway in the brain, which is normally activated by neurons in the ventral tegmental area (VTA) that function, in part, to signal reward. However, DA neurons are not a homogenous population, but rather are composed of multiple subtypes with distinct functions. During the first few years of my postdoctoral fellowship, using a new rabies virus-based transsynaptic tracing approach, I constructed a detailed input-output map of VTA-DA neurons, including the inputs to specific subpopulations of VTA-DA neurons. In a subsequent set of pilot experiments, I conducted a brain-wide input screen using the rabies virus to identify the locus of synaptic changes onto VTA-DA neurons that occur after a single administration of a variety of drugs. Preliminary data from this screen suggest that certain inhibitory inputs (from the globus pallidus external segment, or GPe, and the medial shell of the nucleus accumbens) are altered by a single dose of cocaine, and that these changes occur specifically onto the VTA-DA neurons projecting to the lateral shell of the nucleus accumbens. In addition, these changes occur in animals that display behavioral sensitization, suggesting that these inputs may be specifically modified in this aspect of addictive behaviors. To more rigorously understand the roles of these inputs in addiction, I propose to analyze the roles of these synaptic adaptations in a self-administration model of addiction. First, I will examine the nature of the plasticity occurrng from these defined synapses onto VTA-DA neurons. Secondly, I will use the knowledge of these changes to direct optogenetic and chemogenetic manipulations in-vivo in an attempt to prevent or reverse the drug- induced behavioral alterations occurring with cocaine self-administration. Lastly, in the independent phase, I will use the self-administration model, but now focus on the role of lateral habenula inputs to the midbrain for the development and expression of cocaine withdrawal. As the LHb contains unique subtypes that project either to VTA-DA neurons or GABA neurons in the rostrotegmental area (RMTg), I will analyze the effects of self-administration and withdrawal on the LHb?VTA-DA and LHb?RMTg-GABA connections, the role that each input plays in withdrawal, and if manipulations of each subtype can either exaggerate or alleviate symptoms of withdrawal. As my preliminary data suggest that the rabies virus can be used as a non-biased technique to screen for drug-induced synaptic alterations, the immense power of this technique in generating hypotheses will be highly useful in generating a unique research plan independent from that of my postdoctoral mentor.